Replication analysis of myopia-associated genes in Japanese cohort and in highly myopic patients using genome-wide association study

Abstract

Purpose: Recently, two large consortium for myopia, Consortium for Refractive Error and Myopia (CREAM) and 23andMe, reported the results of genome-wide association analysis (GWAS) on spherical equivalent (SE) refractive error and myopia age of onset and found that 29 SNPs and 35 SNPs were associated with myopia in Caucasians, respectively. In this study, we analyzed the associations of these myopia-associated genetic loci on axial length (AL) and SE in Japanese.

Methods: We included 3,248 Japanese healthy volunteers from the Nagahama Study and 500 Japanese unrelated highly myopic (AL≧26mm) patients from Kyoto University Hospital. From the results of above two GWAS, 50 genes were seemed to have associations with myopia. To investigate and replicate these associations in Japanese, we conducted 2 quantitative trait locus (QTL) analyses using the Nagahama cohort and 1 GWAS using both cohorts. The QTL analyses were conducted on AL and SE and GWAS was conducted on the existence of high myipia. We genotyped 3,248 healthy volunteers using either Illumina OmniExpress2.5M or HumanHap610K, and GWAS was conducted on the existence of high myopia using either HumanHap550K or HumanHap660K. For each analyses, VEGAS (Versatile Gene-based Association Study) program were applied to perform gene-based association tests on myopia-related 50 genes.

Results: In our 2 QTL analyses and 1 GWAS, only 4 genes (RASGRF1, BMP4, GJD2, and CACNA1D) showed statistical significance in all three analyses and 7 genes (B4GALNT2, SH3GL2, SETMAR, ADAMTSL1, BICC1, SFRP1, and TOX) showed statistical significance in two of the three analyses, whereas 27 genes did not show significance in any of our study. In GWAS on high myopia, 17 genes showed statistical significance. However, only 6 genes showed statistical significance in QTL analyses on AL and SE among these 17 genes, respectively. On the other hand, QTL analyses on AL and SE showed statistically significant 11 genes each, and as many as 8 genes had statistical significance among these 11 genes in common.

Conclusions: For previously reported myopia-related genes, we showed the results of gene-based tests on AL, SE, and existence of high myopia in our study cohort in Japan. Our data suggests that the genetic background of Caucasians and that of Japanese for myopia, and that of myopia and high myopia, could be different.

Keywords: 539 genetics • 605 myopia